Thursday, November 01, 2007

Teleology and Indeterminacy

Yesterday, at the LOGOS Seminar, Manolo M presented two ideas for responding to Fodor on teleological/functional solutions to the “Disjunction Problem.”

I haven’t (re-?)read Fodor’s stuff yet, but if I followed correctly, Fodor's general point was that there arguably are pairs of distinct properties such that, nonetheless, there is no fact of the matter as to whether a given mental state has the function of signalling one as opposed to the other. (As Oscar remarked, plausible examples might be harder to come with if a restriction to natural (enough) properties is in place.)

This sounds right. But, as Sebas also worried, it’s not clear in which sense the resulting indeterminacy is not precisely one the defender of the teleological/functional proposal would independently predict and willingly embrace.

Any views?

2 comments:

Oscar Cabaco said...

Manolo Martínez presented yesterday a well known argument to the effect that Fodor’s teleosemantics leaves content underdetermined. So, for example, the concept employed by a frog to catch flies can be described in teleosemantics as having the content “fly” or as having the content “Small black thing”. After that, Manolo provided two interesting solutions to this problem, but I asked him about a third “partial” solution based on certain thesis defended by David Lewis. If we assume that a better interpretation for a language is that that assigns natural properties to the terms in that language (everything else being equal) then we can avoid having to ascribe the content “small black thing” to the former concept, because being a small black thing is not a natural property (unlike being a fly).

Back then, I assumed that this could only be a partial solution, because it doesn’t preclude cases in which the two possible interpretations assign natural properties to the same concept. For example, in a suitable environment the frog’s concept could be taken to mean “fly” or “insect” (note that both being a fly and being an insect are natural properties), and it’s left undetermined by teleosemantics which is the correct interpretation of this concept in this environment.

But now I think that even if we face two different interpretations that assign different natural properties to the same terms we still have a way to choose one of them. As David Lewis argues, naturalness is a question of degree. There’s no clear cut distinction between natural and non natural properties. Instead, properties or kinds are more or less natural than other proper properties or kinds. (Given the role natural properties are supposed to play, I tend to think that this should be wrong. Nevertheless, I But I will leave aside my own worries and I’ll assume that D. Lewis was right about this.)

An obvious consequence is that it seems fair to rephrase Lewis thesis as follows: if we have two different interpretations for the same language we should prefer (everything else being equal) that interpretations that assigns to the terms in that language properties that are more natural than the properties assigned by the other interpretation to the same terms.

This allow us to choose “fly” as the content of the frog’s concept in the former case because being a fly is more natural than being an insect. (I know that this is a crucial point in my argumentation but I don’t have time to argue for it: this post is already too long. If I remember well, an interesting reading about this topic is Dupre (1981) “Natural Kinds and Biological Taxa”…)

This narrows a lot the underdetermination that is supposed to be a consequence of teleosemantics, because it is far more difficult to find competing interpretations that assign properties to the terms of a language that are “equally natural” or natural in the same degree. In fact, could someone provide an example of two different interpretations of the frog’s concept of “fly” that assign two equally natural properties/kinds to that concept?

Dan López de Sa said...

... being a fly is more natural than being an insect. I know that this is a crucial point in my argumentation but I don’t have time to argue for it...

I'm not sure that this could be sucesfully argued for. But anyway, you general point remains: plausible examples might be harder to come with if a restriction to natural (enough) properties is in place.